Reasoning about causal relationships (e.g., forming and testing hypotheses) is increasingly seen as an important focus of early science education and as a central mechanism of cognitive development. In this proposal I examine the strategies young children use in causal reasoning tasks. In particular, I suggest that two classes of strategies may guide reasoning: those consistent with conceptions of natural causes and those consistent with intentional causes. The primary focus of this proposal will be the reasoning strategies used by five-year-olds. Participants will be presented with stories describing novel causal relations (e.g., a substance that turns blue in the freezer) and asked to make predictions and interpretations about these relations. At least as adults we have different ways of making inductions and judging the certainty of inferences in natural and intentional contexts. Children may have different strategies than adults and/or use their strategies differently. For example, children may believe that more events and behaviors are under voluntary control than do adults. Alternatively, or in addition, children may have a general bias to use the weaker reasoning strategies characteristic of intentional causation. One consequence of this perspective is that children's difficulties with formal scientific reasoning may be less of a general inability than a misunderstanding or a misapplication of strategies appropriate to particular content domains. Further, if strategies really are part of common sense then they are part of the background knowledge that children bring to learning environments. Causal reasoning strategies may be both sources of error and possible frameworks for understanding elaborated scientific content.